mTOR signalling has central control of cell growth and proliferation in all eukaryotes analyzed so far 
. Consistent with mTOR signalling regulating growth, human genetic defects that are associated with upregulated mTOR activity manifest as abnormal cell growth and proliferation. Indeed, many negative regulators of mTOR signalling are known human tumour suppressors. As a result mTOR signalling is currently the most targeted signalling pathway in drug development for the treatment of cancers.
It is known that mTOR signalling is essential for the early developmental programmes of metazoans. For instance Drosophila
or C. elegans
TOR loss-of-function mutations leads to developmental arrest 
and homozygous mTOR−/−
mouse embryos die shortly after implantation due to impaired cell proliferation in both embryonic and extra-embryonic compartments 
. mTOR itself forms the core of two different complexes and of these TOR complex 1 is thought to be responsible for regulating cell growth and proliferation 
. We show that planarian Smed-tor
homologs of TOR
respectively and members of TORC1, are necessary for blastema growth during regeneration ( and Figure S12
). RNAi experiments for Smed-tor
result in a lack of the first mitotic peak and blastema during regeneration and show lower general levels of proliferation through the whole regenerative process. Recently it has been shown that planarian amputation triggers two peaks in neoblast mitoses early in regeneration. The first mitotic peak is a body-wide response to any injury and the second response is induced only when injury results in missing tissue 
. RNAi experiments for Smed-tor
show that these animals lack the first mitotic peak of regeneration and thus lack a response to injury. Our results suggest both that the first mitotic peak is necessary for blastema formation and that this first mitotic peak requires the activation of mTOR signalling. The second mitotic peak is a response to missing tissue and it triggers neoblast differentiation 
. Although reduced, Smed-tor
RNAi planarians did show a second mitotic peak () and some ability to restore missing structures at the wound site within old tissues without making a blastema (). The role of mTOR signalling in regeneration 
and in the regulation of stem cells 
is just starting to be elucidated. Our results support a key role for mTOR signalling in controlling stem cell proliferation and growth during regeneration.
In this study we have also characterised the planarian homolog of hSMG-1
. SMG-1 has widely been linked to NMD 
and SMG-1 has been also shown to have functions independent of NMD, mostly related to cellular stress responses 
. However, its physiological roles are not yet well understood and only few studies have been performed in vivo
. In C. elegans
Smg-1 is dispensable for survival and development while other components of NMD in these organisms are essential 
. In mice, SMG-1 is dispensable for implantation and gastrulation but is critical thereafter for basic differentiation, a phenotype presumed to reflect its function in mediating NMD 
Here we have showed that Smed-smg-1(RNAi) results in higher and extended mitotic responses to injury at 6 hR. Following this levels of proliferation remain raised and this contributes to these animals undergoing continuous growth within the blastema. Cells fail to terminally differentiate, proliferating cells remain in aberrantly high numbers and accumulate inappropriately within the blastema. Thus, Smed-smg-1 in contrast to planarian mTORC1 components is required for restricting the response to amputation and the growth of the blastema. Our data suggest that uncontrolled growth occurs in a gradient along the dorso-ventral axis, increasing from ventral to dorsal regions. Several indications support this observation. In anterior regenerating blastemas, cycling neoblasts accumulate initially in the ventral part of the animal, and this is the only area in which some correct terminal differentiation is observed. In more dorsal blastema regions less differentiation is observed. Instead, we observe the presence of cycling neoblasts invading the dorsal-most part of the brain, a lack of terminally differentiated cells and accumulation of neoblast progeny. This continuous growth of the blastema eventually leads to abnormal and lethal outgrowths, with all the outgrowths initially formed at the dorsal level consisting of cycling neoblasts and progeny. These observations suggest that normal planarian growth may follow a ventral to dorsal pattern. During homeostasis, we observed uncontrolled proliferation that also results in lethal ectopic outgrowths. To investigate the possibility that the phenotype observed during homeostasis is also a response to the injury caused after RNAi injections we performed experiments of RNAi by feeding. However, we also observed that Smed-smg-1 regulates the mitotic response to feeding. Thus we were unable to ascertain whether or not Smed-smg-1 also has a role as a homeostatic brake for growth control or whether it is specific to providing a brake to stimuli that promote growth. Nevertheless, we have uncovered a new role for Smg-1 in stem cell regulation, regeneration and growth. It will be very interesting to investigate the possible evolutionary conservation of these novel roles in other systems.
It has already been shown that planarians provide a model system with which to study tumour suppression and adult stem cell lineages in vivo
. RNAi experiments on the planarian homologs of several known human suppressors such as PTEN or p53 have also shown to lead to outgrowths 
. Although the Smed-smg-1(RNAi)
phenotype shares some characteristics with Smed-PTEN
such as hyper-proliferation, hyperplasia and breakdown of the sub-epithelial basement membrane that surrounds the animal, it also has characteristics not previously described for other planarian RNAi phenotypes. Firstly, we observe the presence of cycling neoblasts in front of the eyes, a region devoid of neoblasts. Secondly we observe hypertrophy of the ventral epidermis.
We describe almost polar opposite roles for planarian mTORC1 and Smed-smg-1 in planarian regeneration. Firstly, mTORC1 and Smed-smg-1 have opposing effects on proliferation and blastema growth (, ). Secondly, we found that while Smed-tor RNAi maintains differentiation at the wound site in the absence of blastema formation, terminal differentiation is impeded in Smed-smg-1 RNAi animals (, ). Combinatorial RNAi experiments showed the Smed-smg-1(RNAi) uncontrolled growth requires mTOR signalling to manifest (). Supporting these results, experiments combining rapamycin and RNAi also showed that rapamycin is able to prevent the outgrowths and increase the survival rate of Smed-smg-1 RNAi planarians in about 50% of animals. While these experiments confirm that loss of mTORC1 generally inhibits growth it also raises the possibility that Smed-smg-1 has either a direct or indirect interaction with mTOR signalling. The broad expression patterns of Smed-smg-1, Smed-tor and Smed-raptor in the planarian body, with similar distributions of expression in neoblasts, progeny and differentiated cells also suggests this is a possibility. Future genetic and biochemical work in cell lines and other models systems will help elucidate the relationship between SMG-1 and mTORC1 signalling.
Previously published data showed that rapamycin does not affect basal proliferation in planarians 
. However, only homeostatic planarians were tested injecting doses of 20 nM rapamycin. We also found that 20 nM does not affect proliferation during homeostasis. However, already 20 nM rapamycin was able to significantly decrease proliferation in regenerating planarians while 30 and 40 nM rapamycin doses reduced proliferation in both homeostatic and regenerating planarians. We observed differences between the rapamycin treatment and the inhibition of mTORC1 in planarians. Rapamycin was able to decrease the general levels of planarian proliferation without abolishing the mitotic response to injury and did not affect blastema formation and growth. Significantly, rapamycin does not completely inhibit mTORC1 activity in mammalian cells and clearly affects some TORC1 targets more than others, depending on cell type and context 
Altogether our results indicate a novel function for SMG-1. Our findings support a model where SMG-1 acts as break on growth and proliferation. The altered patterns of proliferation and the presence of abnormal invasive cells disrupting the epithelial-mesenchymal interactions we have observed in Smed-smg-1(RNAi)
planarians are hallmarks in the early progression of most human cancers. To date reports from the COSMIC database describe mutations of SMG-1 in human breast cancer cell lines and hSMG-1 RNA is detected only at low levels in lung carcinoma and melanoma cell lines 
. These reports together with our new functional findings indicate SMG-1 is likely to be a potential human tumour suppressor gene product. Given this new physiological role for Smg-1
it will now be important to investigate whether mutations in hSMG-1 results in cell growth and/or invasiveness contributing to the aetiology of cancers.
The defects observed after Smed-smg-1(RNAi)
may not result from defects in the previously characterised role in the NMD pathway. Very recently, a planarian RNAi screen has shown that several components of the NMD pathway in planarians either show no phenotype after RNAi or show a different one to that observed in Smed-smg-1(RNAi)
. This suggests that the phenotypes observed for Smed-smg-1
may be independent of NMD, alternatively our results may reflect unknown roles for a subset of NMD activities requiring SMG-1 but not other NMD components. A precise understanding of the mechanism(s) by which SMG-1 affects impacts growth and proliferation in physiological conditions may be important in the development of novel cancer therapies and other diseases. The identification of a candidate tumour suppressive function for SMG-1 further validates the planarian system as a source of novel insights relevant to human disease processes.